Modify the prescanner to allow compiler directives to appear in
macro expansions, and adjust the parser to accept a semicolon
as a directive terminator.

Differential Revision: https://reviews.llvm.org/D150780

Value::ValueType is a superset of AddressType. Add a function to
convert an AddressType into a Value::ValueType.

Differential Revision: https://reviews.llvm.org/D150826

[RFC][MC][MachO]Only emits compact-unwind format for "canonical" personality symbols. For the rest, use DWARFs.

Details: https://github.com/rust-lang/rust/issues/102754

The MachO format uses 2 bits to encode these personality funtions, with 0 reserved for "no-personality".
This means we can only have up to 3 personality. There are already three popular personalities:  __gxx_personality_v0, __gcc_personality_v0, and __objc_personality_v0.
As a result, any system that needs custom-personality will run into a problem.

This patch implemented jyknight's proposal to simply force DWARFs for all non-canonical personality functions.

Differential Revision: https://reviews.llvm.org/D144999

[RISCV] Reduce dependency on RISCV::RVVBitsPerBlock for calculating vector size for -mrvv-vector-bits.

We can use the minimum value of the BuiltinType's ElementCount and
the element size.

This needs to be done to support LMUL!=1 types anyway.

I did have to make an ordering change in the error checks in
HandleRISCVRVVVectorBitsTypeAttr to check if the type is an RVV
VLS type before checking the size.

Reviewed By: jhuber6

Differential Revision: https://reviews.llvm.org/D150846

This new functionality will help us avoid duplicated code in various
places in the testing infrastructure. Since the string representation
of the wide numbers is to be used by tests, to keep it simple, we
zero-pad the strings.

Reviewed By: lntue

Differential Revision: https://reviews.llvm.org/D150849

Follow up to "Suppress persistent result when running po" (D144044).

This change delays removal of the persistent result until after `Dump` has been called.
In doing so, the persistent result is available for the purpose of getting its object
description.

In the original change, the persistent result removal happens indirectly, by setting
`EvaluateExpressionOptions::SetSuppressPersistentResult`. In practice this has worked,
however this exposed a latent bug in swift-lldb. The subtlety, and the bug, depend on
when the persisteted result variable is removed.

When the result is removed via `SetSuppressPersistentResult`, it happens within the call
to `Target::EvaluateExpression`. That is, by the time the call returns, the persistent
result is already removed.

The issue occurs shortly thereafter, when `ValueObject::Dump` is called, it cannot make
use of the persistent result variable (instead it uses the `ValueObjectConstResult`). In
swift-lldb, this causes an additional expression evaluation to happen. It first tries an
expression that reference `$R0` etc, but that always fails because `$R0` is removed. The
fallback to this failure does work most of the time, but there's at least one bug
involving imported Clang types.

Differential Revision: https://reviews.llvm.org/D150619

Set the charset to UTF-8, link to the actual liscense we used,
claim support for targets LLVM supports instead of listing them
manually, and stop listing individual language standards we support.

This removes the link to Planet Clang as well as the LLVM coverage
reports, as neither has been actively worked on for several years.

We need to call isRVVVLSBuiltinType() before calling getRVVTypeSize().

This is a regression from 6db007a0 that was reported in:
https://github.com/llvm/llvm-project/issues/62697

The assertion was because we require a code synthesis context for the
instantiation of templates, and this reproducer causes a comparison that
doesn't have a parent-template causing one to exists.

This patch fixes it by creating a ConstraintNormalization context.

None of these files are referenced by anything else on the website, and
the content for these files is quite dated or experiments that did not
pan out. I used a few search engines to see if others were linking to
this content (esp the libstdc++ patches) and did not find significant
evidence that they were.

Add user-friendly intrinsic functions for all relaxed SIMD instructions
alongside the existing SIMD128 intrinsic functions in wasm_simd128.h. Test that
the new instrinsics lower to the expected instructions in the existing
cross-project-tests test file.

Reviewed By: aheejin, sbc100

Differential Revision: https://reviews.llvm.org/D150833

This is a workaround to provide a grace period for folks that were
broken by D147089. As a fly-by, also apply comments by Mark I had
somehow missed in the review.

Differential Revision: https://reviews.llvm.org/D150779

  Fixed a test case llvm\test\Object\bigarchive-malformed-header.test fail caused by  windowi OS has 'llvm-ar.exe' instead of llvm-ar.

This reverts commit 3181a6e3.

Summary:

In big archive , there is 32bit global symbol table and 64 bit global symbol table. llvm-ar only support 32bit global symbol table this moment, we need to support the 64 bit global symbol table.

https://www.ibm.com/docs/en/aix/7.2?topic=formats-ar-file-format-big

Global Symbol Tables

Immediately following the member table, the archive file contains two global symbol tables. The first global symbol table locates 32-bit file members that define global symbols; the second global symbol table does the same for 64-bit file members. If the archive has no 32-bit or 64-bit file members, the respective global symbol table is omitted. The strip command can be used to delete one or both global symbol tables from the archive. The fl_gstoff field in the fixed-length header contains the offset to the 32-bit global symbol table, and the fl_gst64off contains the offset to the 64-bit global symbol table.

Reviewers: James Henderson,Stephen Peckham
Differ...

This patch adds the DFP mul and div instructions. This includes both the
double and quad forms of the instructions as well as the record form.

Reviewed By: amyk

Differential Revision: https://reviews.llvm.org/D150787

This patch adds support in the `OpenMPIRBuilder` for generating working
device code for OpenMP target regions. It generates and handles the
result of a call to `__kmpc_target_init()` at the beginning of the
function resulting from outlining each target region, and it also
generates the matching `__kmpc_target_deinit()` call before returning.

It relies on the implementation of target region outlining for host
codegen to handle the production of the new function and the lowering of
its body based on the contents of the associated target region.

Depends on D147172

Differential Revision: https://reviews.llvm.org/D147940

Now that codegen support for zhinx in landed (D149811), we should set
HasLegalHalfType=true for zhinx (see D145071 for the patch doing this
for zfh).

Differential Revision: https://reviews.llvm.org/D150777

Reviewed By: nikic

Differential Revision: https://reviews.llvm.org/D150829

Revert "[clang][AST] Print name instead of type when diagnosing uninitialized subobject in constexpr variables"

This reverts commit 0e167fc0.

This patch causes its assertion to fire when doing AST-dump, so
reverting so that the author has time to fix this.

BSF is not special here and leaving it in default label looked weird.

This reverts commit e13f88d1.

Fix off by one recursion limit check.

This was reduced from the failure that caused the revert in
e13f88d1

[X86][MC] Move the code about fixed register encoding optimization to X86EncodingOptimization.cpp, NFCI

For some reason the inliner calls simplifyInstruction with disembodied
instructions. I consider this to be an API defect. Either the instruction
should always be inserted prior to simplification, or we at least
should pass in the new function for the context.

Differential Revision: https://reviews.llvm.org/D150854

These functions will be used as part of the HLFIR lowering for
forall/where. The contents of the API were requested by @jeanPerier.

The API is designed around that use case, assuming that the caller knows
through some side channel what size to allocate for boxes returned from
the pop() function.

Differential Revision: https://reviews.llvm.org/D150050

Implement hlfir.elemental inlining as proposed in
flang/docs/HighLevelFIR.md.

This is a separate pass to make the code easier to understand. One
alternative would have been to modify the hlfir.elemental lowering in
the HLFIR bufferization pass.

Currently, a hlfir.elemental can only be inlined once; if there are
more uses, the existing bufferization is used instead.

Usage of mlir::applyPatternsAndFoldGreedily was suggested by @jeanPerier

Differential Revision: https://reviews.llvm.org/D149258

This patch enables the tail-folding of simple loops by default
when targeting the neoverse-v1 CPU. Simple loops exclude those
with recurrences or reductions or loops that are reversed.

New tests have been added here:

Transforms/LoopVectorize/AArch64/sve-tail-folding-option.ll

In terms of SPEC2017 only one benchmark is really affected when
building with "-Ofast -mcpu=neoverse-v1 -flto", which is
(+ faster, - slower):

525.x264: +7.0%

Differential Revision: https://reviews.llvm.org/D130618

Extra tests for D148841.

This is a follow-up for c13ed1cc

Documentation is made more exact, term "constraint" is removed entirely,
description of checker option is corrected.

Reviewed By: Szelethus, gamesh411

Differential Revision: https://reviews.llvm.org/D149447

Without this patch SelectionDAG silently drops dbg.values using `!{}` operands.

Related to https://discourse.llvm.org/t/auto-undef-debug-uses-of-a-deleted-value

This causes assignment-tracking to behaviour to match non-assignment-tracking
behaviour after a recent change (see D140990).

Reviewed By: jmorse

Differential Revision: https://reviews.llvm.org/D150767

This is preferred because it does not rely on accurate kill flags.

Differential Revision: https://reviews.llvm.org/D150673